A method and system for staging diabetic kidney disease using deep learning

Abstract

Embodiments herein disclose a method and system for staging diabetic kidney disease using deep learning techniques. An image capturing unit captures a set of ophthalmic images of a user. The ophthalmic images set undergoes pre-processing before being fed to a first deep learning module. The first deep learning module extracts pathological data indicative of vascular abnormalities from the pre-processed set of ophthalmic images. The first deep learning module quantifies the extracted pathological data, and maps them to a stage of diabetic retinopathy and urine protein levels. A second deep learning module receives as input the quantified pathological data, the mapped diabetic retinopathy stage and urine protein levels, and clinical and demographic parameters. Based on this input, the second deep learning module predicts a stage of diabetic kidney disease.

1 . A system comprising:
 an image capturing unit configured to capture a set of ophthalmic images of a person; and   an image processing unit configured to process the set of ophthalmic images as obtained from the image capturing unit,   wherein the image processing unit comprises:
 a pre-processing module configured to pre-process the set of ophthalmic images; 
 a first deep learning module configured to:
 extract pathological data indicative of vascular abnormalities from the pre-processed set of ophthalmic images; 
 quantify the extracted pathological data based on the vascular abnormalities; and 
 map the quantified pathological data to a stage of diabetic retinopathy and urine protein levels; and 
 
 a second deep learning module configured to:
 receive clinical parameters and demographic parameters; 
 receive, from the first deep leaning module, the quantified pathological data, the stage of diabetic retinopathy, and the urine protein levels; and 
 predict a stage of diabetic kidney disease based on: the quantified pathological data, the stage of diabetic retinopathy, the urine protein levels, the clinical parameters and the demographic parameters. 
 
   
     
     
         2 . The system as claimed in  claim 1 , wherein the vascular abnormalities in the set of ophthalmic images is representative of vascular abnormalities in the kidney that leads to leakage of protein in the urine. 
     
     
         3 . The system as claimed in  claim 2 , wherein the image processing unit is configured to divide each image, in the set of ophthalmic images, into four quadrants, wherein based on the vascular abnormalities in each quadrant, the first deep learning module uses at least one deep learning technique to extract and quantify the pathological data in each quadrant. 
     
     
         4 . The system as claimed in  claim 3  wherein the first deep learning module maps the quantified pathological data in each quadrant to urine protein levels indicating the extent of protein leakage in the urine. 
     
     
         5 . The system as claimed in  claim 1 , wherein the pre-processing module is configured to:
 apply Gaussian blur to perform at least one of the following:
 smooth the set of ophthalmic images; or 
 smooth the clinical and demographic parameters, and eliminate noise in the clinical and demographic parameters; and 
   apply Ben Graham pre-processing to at least one of the following:
 the smooth set of ophthalmic images; or 
 the smooth clinical and demographic parameters for de-noising. 
   
     
     
         6 . The system as claimed in  claim 3 , wherein the first deep learning module includes:
 a first trained deep learning model for extracting and quantifying the pathological data in each quadrant of an ophthalmic image in the set of ophthalmic images;   a second trained deep learning model for mapping the quantified pathological data to the stage of diabetic retinopathy; and   a third trained deep learning model for mapping the quantified pathological data to the urine protein levels.   
     
     
         7 . The system as claimed in  claim 6 , wherein the second deep learning module includes a fourth trained deep learning model that:
 receives, from the second and third trained deep learning models, the stage of diabetic retinopathy and the urine protein levels, respectively; and   predicts the stage of diabetic kidney disease based on the stage of diabetic retinopathy and the urine protein levels.   
     
     
         8 . The system as claimed in  claim 1 , wherein the clinical and demographic parameters include at least one of: age, gender, other comorbidities, duration of diabetes, or history of hypertension. 
     
     
         9 . The system as claimed in  claim 1 , wherein the vascular abnormalities in the set of ophthalmic images is representative of at least one of the following: no abnormalities, microaneurysms, dot hemorrhages, blot hemorrhages, hard exudates, cotton wool spots, intraretinal hemorrhages, venous beading, intraretinal microvascular abnormalities, neovascularization, vitreous hemorrhage or preretinal hemorrhage. 
     
     
         10 . The system as claimed in  claim 9 , wherein the stage of diabetic retinopathy is “no diabetic retinopathy” if the vascular abnormalities patterns are representative of no abnormalities. 
     
     
         11 . The system as claimed in  claim 9 , wherein the stage of diabetic retinopathy is mild non-proliferative diabetic retinopathy if the vascular abnormalities patterns is representative of microaneurysms. 
     
     
         12 . The system as claimed in  claim 9 , wherein the stage of diabetic retinopathy is moderate non-proliferative diabetic retinopathy if the vascular abnormalities patterns is representative of microaneurysms, dot hemorrhages, blot hemorrhages, hard exudates, and cotton wool spots. 
     
     
         13 . The system as claimed in  claim 9 , wherein the stage of diabetic retinopathy is severe non-proliferative diabetic retinopathy if the vascular abnormalities patterns is representative of microaneurysms, dot hemorrhages, blot hemorrhages, hard exudates, cotton wool spots, intraretinal hemorrhages, venous beading, and intraretinal microvascular abnormalities. 
     
     
         14 . The system as claimed in  claim 9 , wherein the stage of diabetic retinopathy is proliferative diabetic retinopathy if the vascular abnormalities patterns is representative of microaneurysms, dot hemorrhages, blot hemorrhages, hard exudates, cotton wool spots, intraretinal hemorrhages, venous beading, intraretinal microvascular abnormalities, neovascularization, vitreous hemorrhage or preretinal hemorrhage. 
     
     
         15 . The system as claimed in  claim 1 , wherein the urine protein levels are categorized as: normal, microalbuminuria, or macroalbuminuria. 
     
     
         16 . The system as claimed in  claim 1 , wherein the predicted stage of diabetic kidney disease is classified as one of: no diabetic kidney disease, early stage diabetic kidney disease, advanced diabetic kidney disease, or late stage diabetic kidney disease. 
     
     
         17 . The system as claimed in  claim 1 , wherein the predicted stage of diabetic kidney disease is indicative of the progression of renal failure, wherein the renal failure is categorized as one of: stable, rapid, or slow. 
     
     
         18 . The system as claimed in  claim 1 , wherein the set of ophthalmic images and the clinical and demographic parameters are respectively used as independent input information to the second deep learning module. 
     
     
         19 . The system as claimed in  claim 1 , wherein the set of ophthalmic images includes fundus images of right and left eyes. 
     
     
         20 . The system as claimed in  claim 1 , wherein the prediction by the second deep learning module is representative of a referable criteria to a nephrologist. 
     
     
         21 . The system as claimed in  claim 1 , comprising a display with a user interface, wherein the quantified pathological data, the stage of diabetic retinopathy, the urine protein levels, and the predicted stage of diabetic kidney disease are displayed on the user interface. 
     
     
         22 . The system as claimed in  claim 1 , wherein the image capturing unit is at least one of: a fundus camera or an optical coherence tomography (OCT) machine. 
     
     
         23 . The system as claimed in  claim 1 , wherein the set of ophthalmic images are infrared images. 
     
     
         24 - 47 . (canceled)